INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hub

-0.81

Category

-0.78

 Mans

-0.70

Mand

-0.67

WARE

-0.65

::::

-0.65

faces

-0.62

Henry

-0.62

 Sans

-0.62

Children

-0.61

POSITIVE LOGITS

etheless

0.74

staking

0.74

 disg

0.73

cles

0.72

 Glas

0.70

 depos

0.69

senal

0.69

tsky

0.69

 Cind

0.68

 Vaugh

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.